The proposed project will evaluate the effects of eight major types of alcohol control policies in multiple states for their effects on alcohol consumption and a range of violence and injury mortality outcomes. Policies to be examined in depth include: (1) changes in excise tax rates on alcoholic beverages, (2) changes in minimum purchase/drinking age, (3) privatization of retail sales of wine or distilled spirits, (4) sudden changes in "dram shop" server liability exposure, (5) compulsory server training policies, (6) liquor-by-the-drink legislation, (7) zero or .02 BAC limits for young drivers, and (8) lower BAC (i.e.,.08) limits for adult drivers. For each policy, state-level changes occurring between 1968 and 1988 will be identified. Detailed monthly frequencies of mortality by state for the 21-year study period will be collected, stratified by separate categories of mortality (homicide, suicide, auto crash, pedestrian, drowning, fire, falls, other injury). Similarly, monthly data on aggregate sales of beer, wine and distilled spirits will be collected by state, as well as data on a number of theoretically important control variables, such as disposable income, the consumer price index, age structure of the population, ethnicity, and religious preference. The nature of state-specific policy changes will be assessed via surveys of state legislative service bureaus, surveys of state Alcohol Beverage Control Agency officials, and extensive legal research of the relevant statutes and case law. Policy changes will be analyzed within a quasi-experimental time-series design with multiple comparison groups, using a combination of Box-Jenkins intervention analysis and econometric pooled cross-sectional time-series statistical mod ing. Previous studies on alcohol control policies have frequently been limited to one type of policy (e.g., excise tax or minimum drinking age), one or two outcomes (e.g. alcohol consumption or auto crashes), or limited to a single state or small number of replications. The proposed comprehensive project is a major advance over previous work on alcohol control policies, because multiple policies will be analyzed using a series of natural experiments across states and a series of outcome measures over a 21-year time period. The proposed project builds directly on emerging theory concerning the role of macro-societal structures, policies, and conditions in influencing drinking behavior. The opportunity afforded by the proposed study for comparative analyses of multiple policies, each with multiple replications across states, will significantly advance continuing theory development and testing on the role of broader institutional and public structures and practices on drinking patterns and on the violence and injuries frequently associated with alcohol use. Finally, the study is clearly of great practical significance as many states continue to debate and consider these policies. identification of policies that substantially reduce alcohol-related violence and injury morbidity and mortality may provide opportunities for important savings in the extensive health services costs currently incurred to treat trauma victims.